Development and Validation of a Multimodal Fusion Artificial Intelligence Model for Predicting the Efficacy of Neoadjuvant Treatment of Bladder Cancer

NCT06909643 | Completed

• 1 other identifier: SYSKY-2024-738-01
• Study Type: observational
• Target: 469
• Locations: 1 country
• Sites: 1

Brief Summary

This study is a multi-center observational study without interventions, including the construction of an AI predictive model, with retrospective and prospective testing. The study participants are bladder cancer patients who have undergone imaging examinations, been pathologically diagnosed, and received neoadjuvant treatment, with complete clinical and pathological data. The study plans to enroll 130 patients from our center, collecting corresponding imaging images, and gathering clinical and genomic data to build and internally validate a multimodal AI model. The model's generalization and robustness will be tested to explore the association between multimodal data and the efficacy of neoadjuvant treatment for bladder cancer. The aim is to assist clinicians in predicting and evaluating the efficacy of neoadjuvant treatment for bladder cancer, with the goal of improving patient diagnosis, treatment outcomes, and prognosis.

Conditions

Cancer; Neoadjuvant Therapy

Keywords

Artificial Intelligence; Multimodal Fusion; Therapy Prediction; Magnetic Resonance Imaging; Digital Pathology

Outcome Measures

Primary Outcomes (1)

• AUC (Area Under the Receiver Operating Characteristic Curve)

  A comprehensive metric reflecting the overall discriminative ability of the AI model, which integrates the model's sensitivity and specificity across all possible threshold values. It quantifies the probability that the model will correctly rank a randomly selected therapy-sensitive patient higher than a randomly selected therapy-insensitive patient.

  For each enrolled patient, the AI model's prediction results will be generated within several days after neoadjuvant therapy. The AUC of the model will be evaluated upon study completion, an average of 3 years.

Secondary Outcomes (2)

• sensitivity

  For each enrolled patient, the predictive results of AI model will be obtained in several days after neoadjuvant therapy, and the sensitivity of the AI model will be evaluated through study completion, an average of 3 year.

• specificity

  For each enrolled patient, the predictive results of AI model will be obtained in several days after neoadjuvant therapy, and the specificity of the AI model will be evaluated through study completion, an average of 3 year.

Study Arms (1)

Patients with bladder cancer undergoing neoadjuvant therapy

Patients pathological diagnosed with bladder cancer undergoing neoadjuvant therapy.

Diagnostic Test: Artificial intelligence (AI)-based predictive model

Interventions

Artificial intelligence (AI)-based predictive model DIAGNOSTIC_TEST

Collect magnetic resonance imaging and pathological slides of resected tumor of the enrolled patients. Analyze the data using the AI model to generate predictive results (sensitive or insensitive to the neoadjavant therapy). No intervention to patients would be performed in this diagnostic test study.

Patients with bladder cancer undergoing neoadjuvant therapy

Eligibility Criteria

• Sex: all
• Healthy Volunteers: No
• Age Groups: Child (0-17), Adult (18-64), Older Adult (65+)
• Sampling Method: Non-Probability Sample

Study Population

Patients with pathologically confirmed bladder cancer who undergo neoadjuvant therapy and radical cystectomy are planned to be enrolled in this diagnostic test to assess the model's clinical application capability.

You may qualify if:

• Bladder occupying lesions, with histopathological confirmation of bladder cancer after resection.
• Planned neoadjuvant therapy and radical cystectomy.

You may not qualify if:

• Patients who have not undergone standard bladder imaging examinations or have missing imaging or pathological data.
• Patients who have received local treatments (such as interventional embolization) or systemic treatments (such as radiotherapy, chemotherapy, immunotherapy, or targeted therapy).
• Poor quality of imaging or pathological images.

Sponsors & Collaborators

• Sun Yat-Sen Memorial Hospital of Sun Yat-Sen University: lead

Study Sites (1)

Sun Yat-sen Memorial Hospital of Sun Yat-sen University

Guangzhou, Guangdong, 510080, China

Location

Biospecimen

Retention: SAMPLES WITHOUT DNA

Histopathological slides of formalin-fixed, paraffin-embedded tumors resected from patients with bladder cancer undergoing radical tumor resection.

MeSH Terms

Conditions

Neoplasms

Interventions

Artificial Intelligence

Intervention Hierarchy (Ancestors)

Algorithms; Mathematical Concepts

Study Design

• Study Type: observational
• Observational Model: COHORT
• Time Perspective: OTHER
• Sponsor Type: OTHER
• Responsible Party: SPONSOR

Study Record Dates

• First Submitted: March 19, 2025
• First Posted: April 3, 2025
• Study Start: January 1, 2022
• Primary Completion: December 31, 2025
• Study Completion: December 31, 2025
• Last Updated: June 3, 2026

Record last verified: 2025-04

Data Sharing

• IPD Sharing: Will not share

Locations

CN (1)